The scientific data supporting, or not, your position have been posted several pages back. The only one who picked up on them was bigshot and he probably knew the answer anyway. Everyone else, including you, has either ignored them, (true in your case), or chosen to dismiss them as speculative or vague, both expressions being laughable considering no other real world data has been presented to either confirm or refute them.
So in light of the only scientific evidence we have so far, it would appear that human hearing is incapable of discerning any audible effect due to the resonances presented in the average CSD, the timebase isn't long enough by a goodly margin. The CSD's shown as examples in the links you provided don't contradict that position either, if anything they support it.
If I was a loudspeaker or headphone designer would I run CSD's? Yes sure I would, but they would be only one of many measurements I'd make. But to inflate their importance and say they prove something is a reach, with the current knowledge we have. It should also be remembered, in the case of headphones at least, CSD's are anything but consistently repeatable. If the above is true, then any perception of speed or not in a headphone or speaker is due to a frequency response imbalance, all things being equal.
I don't want to put words in his mouth and I'm sure he won't hesitate to correct me if I'm wrong, but it is I think, what bigshot has been trying to get across to you.